The localization problem is a widely known and researched project in sensor networks. Different procedures provide more precise results in different situations. In order to ensure the correctness of an algorithm in a given system in most cases it is inevitable to do tests which are usually costly and slow.
My task was to develop a framework for the company’s sensor network which can be used in localization based on received signal strength measurement. The algorithm of the framework is an exchangeable part of the system to ensure the possibility of using different localization methods and thus provide comparable results.
In my thesis I introduce various tools and methods which I used in localization, the different types of Received Signal Strength Indicator-based (RSSI) localization technology (RSSI distance estimation, fingerprint) and the algorithms used by these methods (multilateration, least squares, SMV, k-NN, etc.). After becoming familiar with the sensor network and studying the relevant professional literature I introduce the realized framework, which includes the configuration of measurements, analysis of the data, the database, the program running on the receivers and the results.
At the end of my thesis I make some recommendations for further development based on the results of some localization measurements.